gbsplay segfault on midi about gbsplay HOT 5 CLOSED

Should be fixed by 711af96

from gbsplay.

Thanks for this !!
It's better but it's not working good. All the files are 1ko (the very first notes).

And i'm not sure about this, but they may not have the correct length and i don't hear the battery channel (4).

from gbsplay.

with the help of nadenislamarre we've made an improvement.
It's about gbs.c from line 116 to 122:
//if (gbs->silence_start &&
// (gbs->ticks - gbs->silence_start) / GBHW_CLOCK >= gbs->silence_timeout) {
//if (gbs->subsong_info[gbs->subsong].len == 0) {
// gbs->subsong_info[gbs->subsong].len = gbs->ticks * GBS_LEN_DIV / GBHW_CLOCK;
//}
//return gbs_nextsubsong(gbs);
//}

then this command work: ~/dev/gbsplay$ ./gbsplay -o midi ~/Téléchargements//Smurfs.gbs
and return all the midi files with the correct lenght.

But there still a problem with the channels, the velocity, and others things that i don't know the english music word to explain well.

But, the music is exactly the good one with this command:
~/dev/gbsplay$ ./gbsplay -o alsa ~/Téléchargements//Smurfs.gbs

Why so much differences ? Specially the channel one.

from gbsplay.

I've never used the midi output plugin myself, but from the code it looks like it ignores channel 4, so that part is expected.

from gbsplay.

Not using channel 4 makes kind of sense: In a MIDI drum track (normally on channel 10), every note stands for one drum sound (35 = bass drum 1, 39 = hand clap, 46 = open hi-hat, etc.).
It's not so easy to calculate the current drum sound from current the noise and envelope on channel 4.

If somebody wanted to do some real cool programming, a conversion could propably be built, but I wouldn't even want to try…

from gbsplay.